The airborne contagiousness of respiratory viruses: A comparative analysis and implications for mitigation

Abstract

The infectious emission rate is a fundamental input parameter for airborne transmission risk assessment, but data are limited due to reliance on estimates from chance superspreading events. This study assesses the strength of a predictive estimation approach developed by the authors for SARS-CoV-2 and uses novel estimates to compare the contagiousness of respiratory pathogens. We applied the approach to SARS-CoV-1, SARS-CoV-2, MERS, measles virus, adenovirus, rhinovirus, coxsackievirus, seasonal influenza virus and Mycobacterium tuberculosis (TB) and compared quanta emission rate (ER_q) estimates to literature values. We calculated infection risk in a prototypical classroom and barracks to assess the relative ability of ventilation to mitigate airborne transmission. Our median standing and speaking ER_q estimate for SARS-CoV-2 (2.7 quanta h^-1) is similar to active, untreated TB (3.1 quanta h^-1), higher than seasonal influenza (0.17 quanta h^-1), and lower than measles virus (15 quanta h^-1). We calculated event reproduction numbers above 1 for SARS-CoV-2, measles virus, and untreated TB in both the classroom and barracks for an activity level of standing and speaking at low, medium and high ventilation rates of 2.3, 6.6 and 14 L per second per person (L s^-1 p^-1), respectively. Our predictive ER_q estimates are consistent with the range of values reported over decades of research. In congregate settings, current ventilation standards are unlikely to control the spread of viruses with upper quartile ER_q values above 10 quanta h^-1, such as SARS-CoV-2, indicating the need for additional control measures.

Keywords: Influenza; Measles; Quanta emission rate; SARS-CoV-2 (COVID-19) airborne transmission; Ventilation.

Graphical abstract

Fig. 1

Predictive ER_q distributions for the standing and speaking activity level. Boxes span the interquartile ranges, with whiskers extending from the 5th–95th percentile values and the median denoted by the vertical line in each box.

Fig. 2

Predictive cumulative ER_q distributions for a tuberculosis (TB) ward with up to six patients based on the resting, oral breathing expiration and bacillary load (see Supplementary data text) representative of active, untreated TB (Fig. 2a) and after two weeks of treatment (Fig. 2b). Boxes span the interquartile ranges, with whiskers extending from the 5th–95th percentile values and the median denoted by the horizontal line in each box. Labeled gridlines present the cumulative ER_q estimates from seminal human-to-guinea pig transmission trials as follows: 0.72 quanta h⁻¹ from a six-bed ward (after Riley et al., 1959), 1.25 quanta h⁻¹ from a five-bed ward (after Riley et al., 1962), 34 quanta h⁻¹ from a six-bed ward with patients wearing masks (after Dharmadhikari et al., 2012), 43 quanta h⁻¹ from an eight-bed ward with 66% occupancy (after Escombe et al., 2007, calculated as the mean reported individual patient ER_q of 8.2 quanta h⁻¹ times the mean bed occupancy of 5.3 beds), and 138 quanta h⁻¹ from a six-bed ward with no mask use (after Dharmadhikari et al., 2012).

Fig. 3

Individual risk (R, %) and event reproduction numbers (R_event) for the (a) classroom and (b) barracks modeling scenarios. Circles depict results for resting, oral breathing and triangles depict results for standing, speaking. White, gray, and black symbol fill corresponds to the high, medium, and low ventilation rates, respectively.